Pagefind
Tool, static site, by CloudCannon
Assessment
The search for a static site: run it over the output, ship the folder it writes, and nothing is sent anywhere when someone searches. Its lazy-loaded index makes it practical even for thousands of pages.
2026-10-09
Strengths
- Full-text search for a static site with no server and no third party
- Indexes the built HTML after any generator; loads only the index chunks a query needs
- Filters and a drop-in search UI
Limitations
- Index built at deploy time: new content is not searchable until the next build
- The UI's defaults need a little CSS to match a site
Pagefind is a search tool that runs after a static site build: it reads the HTML in the output folder, writes a compressed index split into small chunks, and provides a JavaScript client that fetches only the chunks a query touches.[1] The result is full-text search with no search server and no outside service.
What it is for
Any static site that needs search without a backend. It is generator-agnostic: Eleventy, Hugo, Astro and hand-written HTML all work, because it only sees the output.
Things to know
- Mark the searchable region with
data-pagefind-body, or every navigation bar is indexed too. - Filters come from
data-pagefind-filterattributes in the HTML; a kind or category becomes a facet for free. - The bundled UI can be driven from an input elsewhere on the page with
triggerSearch, which is how a header search box can open a results panel without a search page.[2] - The index folder must be served as static files, including its binary chunks; a server that compresses or rewrites unknown types will break it.
History
- 2026-10-09: article written and published.
- 2026-10-09: entry created.
Comments
Public comments on each entry are coming. Nothing is collected here yet.